STAT 154 Lecture 14: LOOCV

• School: Berkeley
• Department: Statistics
• Course: STAT 154
• Professor: John Tuhao Chen

Document Summary

Short cut formula , linear regression cvcni-tn. ie c. 7cxj - is - and xi fi y. C ven , estimated with whole data th e cy i. = # t ) ( x" x ) = nice , yet , the ) pi. "t yt l x" a , xu , x" the } e - East c nice , act , ) East china " x" a , 7 it ( x. { 1kt let lieu be the residual data , we have nice , from using the ( ent ) ht. Nice , t leaving the t th column "s observation. V , the variation of each data point in redeemed by. Icid the yi ji in is not the same as in use s. Loo w training set y i in by its meaning. C von , th e ly i iii.